NO. 6 CATERPILLAR AND BUTTERFLY — SNODGRASS 25 



The larval tracheal system is highly elaborate (fig. 13), but in no 

 way specialized. The nervous system of a noctuid caterpillar is de- 

 scribed by DuPorte (1914) and that of a hepialid by Swaine (1920- 

 21). In the hepialid Sthenopis thule and an eriocraniid, according to 

 Swaine, there are eight abdominal ganglia, but the connectives in both 

 the thorax and the abdomen are fused into single interganglionic 

 strands. The same is true of a species of Hepialus examined by the 

 writer. The union of the nerve connections is thus a specialized 

 feature in this otherwise relatively primitive group. On the other 

 hand, in caterpillars of the higher Lepidoptera the last two ganglia 

 are united in the seventh abdominal segment. 



The theory that would explain the different forms of holometab- 

 olous larvae on the assumption that they have hatched at different 

 stages of embryonic development assigns the caterpillar to the polypod 

 stage of embryogeny. The polypod stage is that in which rudiments 

 of the gnathal and thoracic appendages are well formed, and vestiges 

 of limbs are present on the abdomen. This stage of the embryo 

 presumably represents a disappearing "centipede" stage in the evolu- 

 tion of the insects. 



The polypod theory as applied to the caterpillar encounters two 

 principal objections. First, it assumes that the abdominal prolegs of 

 the larva are serially homologous with the thoracic legs, an assumption 

 that is highly improbable. Second, if we look at an embryo in the 

 polypod stage of development, such as that shown by Eastham (1930) 

 for Pieris rapae (fig. 14), it is seen that the head is still in a very 

 primitive state. The gnathal segments are not yet added to the 

 embryonic head (emH), and the prospective mouth parts (Md, iMx, 

 2Mx) are mere undifferentiated lobes. Yet, when the larva leaves the 

 egg it is already a perfectly formed young caterpillar. The head is a 

 syncephalon, the mouth parts are fully developed for feeding and 

 silk spinning. All this is quite at variance with the idea that the 

 caterpillar is an arrested polypod stage of the embryo. 



While in most ways the caterpillar is structurally and functionally 

 subservient to the adult, in one respect it has been self-centered. This 

 is in its production and spinning of silk, which seems to have no 

 primary relation to the welfare of the moth or butterfly. The silk is 

 formed in the labial glands, which ordinarily in other insects secrete 

 saliva. Since caterpillars use their silk in various ways, it is not clear 

 what use it may have been to them in the first place. 



True silk is formed only by arthropods, particularly the insects and 

 spiders, but the silk-producing organs may be glands in quite different 

 parts of the body having no homology with one another. Silk produc- 



